bge-reranker-v2-m3 and Prompt-Guard-86M are both text-classification models. See each entry for specifics.
BGE-Reranker-v2-M3 is BAAI's multilingual cross-encoder reranker built on XLM-RoBERTa, designed for re-ranking retrieved passages in multilingual RAG or search pipelines. It jointly encodes query-passage pairs to produce relevance scores, providing higher accuracy than bi-encoder similarity for the same candidate set. Apache 2.0 licensed with text-embeddings-inference support.
As a llama-based compact model, Prompt-Guard-86M focuses on text classification. Weighing in near 86M parameters, Prompt-Guard-86M trades some ceiling for cheaper, faster inference. Prompt-Guard-86M is subject to Llama 3.1 Community terms, so confirm licensing before commercial use. Prompt-Guard-86M ships without a hosted SLA, so budget for self-managed deployment and monitoring.
Pick based on your compute budget and specific task requirements.